CN112433495B - Rehabilitation robot rapid finite time control based on SCN man-machine uncertain model - Google Patents
Rehabilitation robot rapid finite time control based on SCN man-machine uncertain model Download PDFInfo
- Publication number
- CN112433495B CN112433495B CN202011357061.5A CN202011357061A CN112433495B CN 112433495 B CN112433495 B CN 112433495B CN 202011357061 A CN202011357061 A CN 202011357061A CN 112433495 B CN112433495 B CN 112433495B
- Authority
- CN
- China
- Prior art keywords
- human
- robot
- machine
- uncertainty
- finite time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 claims abstract description 54
- 230000006641 stabilisation Effects 0.000 claims abstract description 12
- 238000011105 stabilization Methods 0.000 claims abstract description 12
- 230000005484 gravity Effects 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 6
- 230000003044 adaptive effect Effects 0.000 claims description 4
- 230000000694 effects Effects 0.000 description 4
- 230000001052 transient effect Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 230000002401 inhibitory effect Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000000474 nursing effect Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/04—Programme control other than numerical control, i.e. in sequence controllers or logic controllers
- G05B19/042—Programme control other than numerical control, i.e. in sequence controllers or logic controllers using digital processors
- G05B19/0423—Input/output
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/20—Pc systems
- G05B2219/25—Pc structure of the system
- G05B2219/25257—Microcontroller
Abstract
The invention discloses a rehabilitation robot rapid limited time control method based on a random configuration network (Stochastic Configuration Networks, SCN) man-machine uncertain model. The method is characterized in that: based on a dynamic model of the rehabilitation walking training robot, separating uncertain items of a human-machine system caused by a rehabilitee, and establishing the dynamic model of the rehabilitation walking training robot with human-machine uncertainty; a network estimation model of the human-computer system uncertainty is built based on an SCN method, a motion trail and a speed are taken as network inputs, and the human-computer system uncertainty estimation is obtained by continuously and randomly configuring hidden layer node parameters; the tracking controller is designed based on a finite time stabilization theory, so that the influence of human-machine uncertainty on the tracking precision of the system is restrained, and meanwhile, the robot is ensured to be at any initial position, so that the tracking error system can be stabilized for a finite time.
Description
Technical field:
the invention relates to a control method of a wheeled rehabilitation robot, in particular to the field of control of rehabilitation robots with human-machine uncertainty.
The background technology is as follows:
traffic accidents and population aging increase patients with walking disorders year by year, and the patients with walking disorders cannot get timely and effective exercise training due to the lack of professional rehabilitation staff in China, so that the walking functions are gradually lost, and daily independent life cannot be realized. Along with the application of the rehabilitation walking robot in places such as rehabilitation centers, nursing homes and the like, the problem of shortage of rehabilitation personnel is effectively solved. However, in practical application, human-machine uncertainty is generated when a person contacts with a robot, so that the walking robot is seriously interfered with the tracking of the designated motion trail of a rehabilitation doctor, an ideal rehabilitation effect cannot be achieved, and the robot collides with surrounding objects due to an excessive tracking error, so that the safety of a trainer is threatened. Therefore, the control method of the walking robot is researched, the problem of uncertain contact between the human and the machine is solved, and the method has important significance for guaranteeing the rehabilitation effect and safety of a trainer.
In recent years, trace tracking control of rehabilitation walking robots has many research results, but these results cannot solve the problem of human-machine uncertainty generated in cooperative motion of people and robots and the problem of low transient tracking performance of robots at any initial positions. If the human-machine contact and human-machine system transient performance are not solved, a large tracking error can be generated to threaten the safety of a trainer. To date, there is no fast finite time control method for human-machine uncertainty estimation and any initial position that accommodates human-machine contact uncertainty changes. The invention provides an estimation method for uncertainty of human-machine contact based on a new view angle, researches a rapid limited time control method for compensating human-machine uncertainty, and has important significance for guaranteeing rehabilitation effect and safety of trainers.
The invention comprises the following steps:
the invention aims to:
in order to solve the problems, the invention provides a rapid limited time control method of a rehabilitation robot based on an SCN man-machine uncertain model.
The technical scheme is as follows:
the invention is realized by the following technical scheme:
a rehabilitation robot rapid limited time control method based on an SCN (Stochastic Configuration Networks, SCN) man-machine uncertain model is characterized in that:
1) Based on a dynamic model of the rehabilitation walking training robot, separating uncertain items of a human-machine system caused by a rehabilitee, and establishing the dynamic model of the rehabilitation walking training robot with human-machine uncertainty;
2) A network estimation model of the human-computer system uncertainty is built based on an SCN method, a motion trail and a speed are taken as network inputs, and the human-computer system uncertainty estimation is obtained by continuously and randomly configuring hidden layer node parameters;
3) The tracking controller is designed based on a finite time stabilization theory, so that the influence of human-machine uncertainty on the tracking precision of the system is restrained, and meanwhile, the robot is ensured to be at any initial position, so that the tracking error system can be stabilized for a finite time.
The method comprises the following steps:
step one), based on a dynamics model of a rehabilitation walking training robot, separating uncertain items of a human-computer system caused by a rehabilitation person, and establishing the dynamics model of the rehabilitation walking training robot with human-computer uncertainty, which is characterized in that: the kinetic model of the system is described below
Wherein the method comprises the steps of
M represents robot mass, M represents rehabilitation person mass, r 0 Indicating the distance from the center of the robot to the center of gravity, I 0 The moment of inertia is indicated and the moment of inertia,as coefficient matrix, θ i Representing the horizontal direction and the connection between the robot center and the ith wheel centerIncluded angle between l i Phi is the distance from the center of gravity of the robot to the center of each wheel i Corresponding to each wheel in the horizontal direction i An included angle between the two. u (t) represents a control input force of the robot, f i For the input force of each wheel of the robot, X (t) is the actual motion trajectory of the robot, i=1, 2,3,4.
Separating human-machine uncertainty information caused by a rehabilitee in a model (1), and recordingThe model (1) can be of the following form
Wherein,represents human-machine uncertainty, and
ΔM 0 representation ofSeparated human-machine uncertainty term, ΔB (θ) represents +.>Separated human-machine uncertainty item, M 0 And B (theta) is a coefficient matrix, and L represents the distance from the center of the robot to the center of the wheel.
Order theFrom equation (2) we get the person withRehabilitation walking training robot dynamics model of machine uncertainty:
step two), a network estimation model of the human-computer system uncertainty is built based on an SCN method, a motion track and a speed are taken as network input, and the human-computer system uncertainty estimation is obtained by continuously and randomly configuring hidden layer node parameters, and the method is characterized in that: with the motion trail and speed of the robotAs a network input layer of the SCN, and is connected with the hidden layer through a weight ω and a threshold b, and a gaussian function is used to obtain a hidden layer output G (x (t)).
Wherein the method comprises the steps of
G(x(t))＝[g 1 (ω 1 x(t)+b 1 ),...,g L (ω L x(t)+b L )] T
g j (ω j x(t)+b j ) Output j= (1, 2,..once., L) for the j-th node of the hidden layer, ω h,j Connecting weights of the j-th node of the hidden layer for the h-th input of the input layer, h= (1, 2,., 6), b j Is the threshold of the hidden layer j-th node.
Then, the SCN hidden layer passes the weightsNetwork output connected with output layer to obtain uncertainty estimation of man-machine system>The following are provided:
wherein the method comprises the steps of
And connecting the weight g= (1, 2, 3) of the g output for the j hidden layer node.
Further, according to the human-machine uncertainty estimation error obtained when the node number of the hidden layer is L-1Randomly configuring node parameters of the L hidden layer and enabling the node parameters to meet delta L >0，δ L The expression form is as follows:
due to
Wherein the method comprises the steps of
Therefore, it is
Wherein, parameter 0<r<1,{μ L And is a non-negative real sequence, mu L Is less than or equal to (1-r). When delta L >0, ε L T ε L <(r+μ L )ε L-1 T ε L-1 With random configurationThe number of hidden layer nodes is continuously increased whenEpsilon at this time L T ε L <rε L-1 T ε L-1 Is easy to obtainThe uncertain estimation of the man-machine system can be realized>
Step three) designing a tracking controller based on a finite time stabilization theory, inhibiting the influence of human-machine uncertainty on the tracking precision of the system, ensuring that the tracking error system is stable in a fast finite time at any initial position, and the method is characterized in that: introducing an auxiliary track x a (t) tracking the actual trackThe training track x is designated by doctor d (t) and auxiliary track x a (t) composition, i.e.)> x a (T) satisfy x over a finite time T a (T) =0, and x a (0)＝x 1 (0)-x d (0)。
Wherein the method comprises the steps of
From the expression form of ζ (t), ζ (0) =1, t>At T ζ (T) =0,at the same time, xi (T) is smooth and continuous at the moment T, and can obtain +.>The system trajectory returns to the specified trajectory at a finite time T.
At this time, the track tracking error and the speed tracking error are respectively
Wherein e 1 (t)＝[e 11 (t)，e 12 (t)，e 13 (t)] T ，e 1g (t) the trajectory tracking errors in the x-axis, y-axis and rotation angle directions, respectively. e, e 2 (t)＝[e 21 (t)，e 22 (t)，e 23 (t)] T ，e 2g And (t) represents the speed tracking errors in the x-axis, y-axis and rotation angle directions, respectively.
Design auxiliary variableWherein z is 1 (t)＝[z 11 (t),z 12 (t),z 13 (t)] T ，z 1g (t) auxiliary variable errors in x-axis, y-axis and rotation angle directions, respectively; k (K) 1 ＝diag(k 11 ,k 12 ,k 13 )，/>
Enabling electromechanical system uncertainty estimationWeight matrix->Is beta and
thus can be used forObtain weight errorThe self-adaptive rate of the design weight is as follows
Wherein Γ and K 4 Is an adaptive rate parameter.
The tracking error system obtainable from equation (3) and equation (6) is:
the design finite time controller is as follows:
wherein the method comprises the steps ofK 2 And k 3 Is a controller parameter.
The lyapunov function was established as follows:
deriving (10) along the error system (8)
Substituting the finite time controller (9) into the formula (11) to obtain
Wherein ε=w (t) - β *T G (x (t)) taking the parameter k 3 >And substituting the self-adaption rate (7) into the formula (11) to obtain
According to formula (12), there are further:
thus, from the formula (12)Wherein->μ＝(1+α)/2，1/2<μ<1，/>
Therefore, according to the finite time stabilization theory, the robot can start from any position and the tracking error system can be stabilized for finite time, the rehabilitation walking training robot can rapidly track the motion trail appointed by the doctor in finite time, and meanwhile, the rapid finite adjustment time can be obtained
Step four), based on MSP430 series singlechip provide output PWM signal to motor drive unit, make the robot realize tracking simultaneously to reference track signal's motion track and speed, its characterized in that: taking an MSP430 series singlechip as a main controller, wherein the input of the main controller is connected with a motor speed measuring module, and the output of the main controller is connected with a motor driving module; the motor driving module is connected with the direct current motor; the power supply system supplies power to the respective electrical devices. The control method of the main controller is to read the feedback signal of the motor encoder and the control command signal given by the main controllerAnd->An error signal is calculated. According to the error signal, the main controller calculates the control quantity of the motor according to a preset control algorithm and sends the control quantity to the motor driving module, and the motor rotates to drive the wheels to maintain self balance and move in a specified mode.
The advantages and effects:
the invention relates to a rapid finite time control method for a rehabilitation walking training robot based on SCN estimation man-machine uncertainty, which has the following advantages:
the invention skillfully separates the uncertain items of the human-computer system and establishes a rehabilitation walking training robot dynamics model with human-computer uncertainty; the network estimation model with uncertain man-machine is built based on the SCN method, and the influence of uncertain man-machine on the tracking performance of the system is compensated by designing the fast limited time controller, so that the transient performance of any initial position of the system is improved, and the tracking precision of the system and the safety of a trainer are ensured.
Description of the drawings:
FIG. 1 is a block diagram of the operation of a controller according to the present invention;
FIG. 2 is a graph of a system of the present invention;
FIG. 3 is a schematic diagram of a MSP430 single-chip microcomputer minimal system according to the present invention;
FIG. 4 is a main controller peripheral expansion circuit of the present invention;
fig. 5 is a circuit of the general principles of the hardware of the present invention.
The specific embodiment is as follows:
the present invention will be further described with reference to the accompanying drawings, but the scope of the present invention is not limited by the examples.
A rehabilitation walking training robot rapid limited time control method based on SCN estimation man-machine uncertainty. The method is characterized in that:
1) Based on a dynamic model of the rehabilitation walking training robot, separating uncertain items of a human-machine system caused by a rehabilitee, and establishing the dynamic model of the rehabilitation walking training robot with human-machine uncertainty;
2) A network estimation model of the human-computer system uncertainty is built based on an SCN method, a motion trail and a speed are taken as network inputs, and the human-computer system uncertainty estimation is obtained by continuously and randomly configuring hidden layer node parameters;
3) The tracking controller is designed based on a finite time stabilization theory, so that the influence of human-machine uncertainty on the tracking precision of the system is restrained, and meanwhile, the robot is ensured to be at any initial position, so that the tracking error system can be stabilized for a finite time.
The method comprises the following steps:
step one), based on a dynamics model of a rehabilitation walking training robot, separating uncertain items of a human-computer system caused by a rehabilitation person, and establishing the dynamics model of the rehabilitation walking training robot with human-computer uncertainty, which is characterized in that: the kinetic model of the system is described below
Wherein the method comprises the steps of
M represents robot mass, M represents rehabilitation person mass, r 0 Indicating the distance from the center of the robot to the center of gravity, I 0 Representation turnThe moment of inertia of the material is calculated,as coefficient matrix, θ i Indicating the horizontal direction and the angle between the robot center and the ith wheel center line, l i Phi is the distance from the center of gravity of the robot to the center of each wheel i Corresponding to each wheel in the horizontal direction i An included angle between the two. u (t) represents a control input force of the robot, f i For the input force of each wheel of the robot, X (t) is the actual motion trajectory of the robot, i=1, 2,3,4.
Separating human-machine uncertainty information caused by a rehabilitee in a model (1), and recordingThe model (1) can be of the following form
Wherein,represents human-machine uncertainty, and
ΔM 0 representation ofSeparated human-machine uncertainty term, ΔB (θ) represents +.>Separated human-machine uncertainty item, M 0 B (θ) is a coefficient matrix, L representsDistance from the center of the robot to the center of the wheel.
Order theObtaining a rehabilitation walking training robot dynamics model with human-machine uncertainty from equation (2):
step two), a network estimation model of the human-computer system uncertainty is built based on an SCN method, a motion track and a speed are taken as network input, and the human-computer system uncertainty estimation is obtained by continuously and randomly configuring hidden layer node parameters, and the method is characterized in that: with the motion trail and speed of the robotAs a network input layer of the SCN, and is connected with the hidden layer through a weight ω and a threshold b, and a gaussian function is used to obtain a hidden layer output G (x (t)).
Wherein the method comprises the steps of
b＝[b 1 ,b 2 ,...,b L ] T G(x(t))＝[g 1 (ω 1 x(t)+b 1 ),...,g L (ω L x(t)+b L )] T
g j (ω j x(t)+b j ) Output j= (1, 2,..once., L) for the j-th node of the hidden layer, ω h,j Connecting weights of the j-th node of the hidden layer for the h-th input of the input layer, h= (1, 2,., 6), b j Is the threshold of the hidden layer j-th node.
Then, the SCN hidden layer passes the weightsConnected with the output layer to obtain the network output of the uncertainty estimation of the man-machine systemGo out->The following are provided:
wherein the method comprises the steps of
And connecting the weight g= (1, 2, 3) of the g output for the j hidden layer node.
Further, according to the human-machine uncertainty estimation error obtained when the node number of the hidden layer is L-1Randomly configuring node parameters of the L hidden layer and enabling the node parameters to meet delta L >0，δ L The expression form is as follows:
wherein, parameter 0<r<1,{μ L And is a non-negative real sequence,with increasing number of hidden layer nodes arranged randomly, when +.>Can realize uncertain estimation of a man-machine system +.>Step three), designing a tracking controller based on a finite time stabilization theory, and inhibiting peopleThe influence of uncertainty of the robot on the tracking precision of the system is ensured, and meanwhile, the robot can make the tracking error system stable in a rapid limited time at any initial position, and the method is characterized in that: introducing an auxiliary track x a (t) let the actual tracking track +.>The training track x is designated by doctor d (t) and auxiliary track x a (t) composition, i.e x a (T) satisfy x over a finite time T a (T) =0, and x a (0)＝x 1 (0)-x d (0)。
Wherein the method comprises the steps of
From the expression form of ζ (t), ζ (0) =1, t>At T ζ (T) =0,at the same time, xi (T) is smooth and continuous at the moment T, and can obtain +.>The system trajectory returns to the specified trajectory at a finite time T.
At this time, the track tracking error and the speed tracking error are respectively
Wherein e 1 (t)＝[e 11 (t)，e 12 (t)，e 13 (t)] T ，e 1g (t) the trajectory tracking errors in the x-axis, y-axis and rotation angle directions, respectively. e, e 2 (t)＝[e 21 (t)，e 22 (t)，e 23 (t)] T ，e 2g And (t) represents the speed tracking errors in the x-axis, y-axis and rotation angle directions, respectively.
Design auxiliary variableWherein z is 1 (t)＝[z 11 (t),z 12 (t),z 13 (t)] T ，z 1g (t) auxiliary variable errors in x-axis, y-axis and rotation angle directions, respectively; k (K) 1 ＝diag(k 11 ,k 12 ,k 13 )，Sig(ξ) α ＝[|ξ 1 | α sgn(ξ 1 ),...,|ξ n | α sgn(ξ n )] T ，/>
Enabling electromechanical system uncertainty estimationWeight matrix->Is +.>And is also provided with
Thus, the weight error is obtainedThe self-adaptive rate of the design weight is as follows
Wherein Γ and K 4 Is an adaptive rate parameter.
The tracking error system obtainable from equation (3) and equation (6) is:
the design finite time controller is as follows:
wherein the method comprises the steps ofK 2 And k 3 Is a controller parameter.
The lyapunov function was established as follows:
deriving (10) along the error system (8)
Substituting the finite time controller (9) into the formula (11) to obtain
Wherein ε=w (t) - β *T G (x (t)) taking the parameter k 3 >And substituting the self-adaption rate (7) into the formula (11) to obtain
According to formula (12), there are further:
thus, from the formula (12)Wherein->μ＝(1+α)/2，1/2<μ<1，/>
Therefore, according to the finite time stabilization theory, the robot can start from any position and the tracking error system can be stabilized for finite time, the rehabilitation walking training robot can rapidly track the motion trail appointed by the doctor in finite time, and meanwhile, the rapid finite adjustment time can be obtained
Step four), based on MSP430 series singlechip provide output PWM signal to motor drive unit, make the robot realize tracking simultaneously to reference track signal's motion track and speed, its characterized in that: taking an MSP430 series singlechip as a main controller, wherein the input of the main controller is connected with a motor speed measuring module, and the output of the main controller is connected with a motor driving module; the motor driving module is connected with the direct current motor; the power supply system supplies power to the respective electrical devices. The control method of the main controller is to read the feedback signal of the motor encoder and the control command signal given by the main controllerAnd->An error signal is calculated. According to the error signal, the main controller calculates the control quantity of the motor according to a preset control algorithm, and sends the control quantity to the motor driving module, and the motor rotates to drive the wheels to maintain self flatnessThe scale moves in a specified manner.
The invention skillfully separates uncertain items of a human-computer system caused by a rehabilitee, establishes a dynamic model of a rehabilitation walking training robot with human-computer uncertainty, builds a network estimation model of the human-computer uncertainty based on an SCN method, takes a motion track and speed as network input, obtains the human-computer system uncertainty estimation by continuously and randomly configuring hidden layer node parameters, designs a fast tracking controller based on a finite time stabilization theory, compensates the influence of the human-computer uncertainty on the system tracking precision, ensures that the robot can realize fast finite time tracking of the motion track at any initial position, and can effectively improve the system tracking performance and the safety of a trainer.
Claims (2)
1. A rehabilitation robot rapid limited time control method based on a random configuration network man-machine uncertain model is characterized by comprising the following steps of: based on a dynamic model of the rehabilitation walking training robot, separating uncertain items of a human-machine system caused by a rehabilitee, and establishing the dynamic model of the rehabilitation walking training robot with human-machine uncertainty; constructing a network estimation model of human-machine uncertainty based on a random configuration network method, taking a motion track and a speed as network input, and obtaining human-machine system uncertainty estimation by continuously and randomly configuring hidden layer node parameters; the method has the advantages that the tracking controller is designed based on the finite time stabilization theory, the influence of human-machine uncertainty on the tracking precision of the system is restrained, and meanwhile, the fact that the robot is at any initial position is guaranteed, so that the tracking error system is stable in a quick and finite time manner; the method comprises the following steps:
1) Based on a dynamic model of the rehabilitation walking training robot, separating uncertain items of a human-machine system caused by a rehabilitee, and establishing the dynamic model of the rehabilitation walking training robot with human-machine uncertainty;
2) Constructing a network estimation model of human-machine uncertainty based on a random configuration network method, taking a motion track and a speed as network input, and obtaining human-machine system uncertainty estimation by continuously and randomly configuring hidden layer node parameters;
3) The method has the advantages that the tracking controller is designed based on the finite time stabilization theory, the influence of human-machine uncertainty on the tracking precision of the system is restrained, and meanwhile, the fact that the robot is at any initial position is guaranteed, so that the tracking error system is stable in a quick and finite time manner;
based on the dynamics model of the rehabilitation walking training robot, separating the uncertainty items of a human-machine system caused by a rehabilitee, establishing a dynamics model of the rehabilitation walking training robot with human-machine uncertainty, wherein the dynamics model of the system is described as follows
Wherein the method comprises the steps of
M represents robot mass, M represents rehabilitation person mass, r 0 Indicating the distance from the center of the robot to the center of gravity, I 0 The moment of inertia is indicated and the moment of inertia,as coefficient matrix, θ i Indicating the horizontal direction and the angle between the robot center and the ith wheel center line, l i Phi is the distance from the center of gravity of the robot to the center of each wheel i Corresponding to each wheel in the horizontal direction i An included angle between the two; u (t) represents a control input force of the robot, f i For the input force of each wheel of the robot, X (t) is the actual motion trajectory of the robot, i=1, 2,3,4;
separating dieHuman-computer uncertainty information caused by a rehabilitee in (1), recordThe model (1) can be of the following form
Wherein,represents human-machine uncertainty, and
ΔM 0 representation ofSeparated human-machine uncertainty term, ΔB (θ) represents +.>Separated human-machine uncertainty item, M 0 B (theta) is a coefficient matrix, and L represents the distance from the center of the robot to the center of the wheel;
let x 1 (t)＝X(t),Obtaining a rehabilitation walking training robot dynamics model with human-machine uncertainty from equation (2):
a network estimation model of human-machine uncertainty is built based on a random configuration network method, a motion trail and speed are taken as network input, the uncertainty estimation of a human-machine system is obtained by continuously and randomly configuring hidden layer node parameters, and the motion trail and speed of a robot are taken as network inputAs a network input layer of a random configuration network, connecting with an hidden layer through a weight omega and a threshold b, and obtaining hidden layer output G (x (t)) by using a Gaussian function;
wherein the method comprises the steps of
b＝[b 1 ,b 2 ,...,b L ] T G(x(t))＝[g 1 (ω 1 x(t)+b 1 ),...,g L (ω L x(t)+b L )] T
g j (ω j x(t)+b j ) Output j=1, 2 for the j-th node of the hidden layer hj Connecting the weight of the j-th node of the hidden layer for the h-th input of the input layer, wherein h=1, 2, …, b j A threshold value for the j-th node of the hidden layer;
then, randomly configuring the network hidden layer passing weightNetwork output connected with output layer to obtain uncertainty estimation of man-machine system>The following are provided:
wherein the method comprises the steps of
Connecting the weight g=1, 2,3 of the g output for the j hidden layer node;
further, according to the human-machine uncertainty estimation error obtained when the node number of the hidden layer is L-1Randomly configuring node parameters of the L hidden layer and enabling the node parameters to meet delta L ＞0，δ L The expression form is as follows:
wherein, the parameter 0 is less than r and less than 1, { mu } L And is a non-negative real sequence, mu L ≤1-r,With increasing number of hidden layer nodes arranged randomly, when +.>Can realize uncertain estimation of a man-machine system +.>
The tracking controller is designed based on a finite time stabilization theory, the influence of human-machine uncertainty on the tracking precision of the system is restrained, meanwhile, the robot is ensured to be at any initial position, a tracking error system can be enabled to be stable for a finite time, and an auxiliary track x is introduced a (t) tracking the actual trackThe training track x is designated by doctor d (t) and auxiliary track x a (t) composition, i.e x a (T) satisfy x over a finite time T a (T) =0, and x a (0)＝x 1 (0)-x d (0)；
Wherein the method comprises the steps of
From the expression form of ζ (T), it can be seen that when ζ (0) =1, T > T, ζ (T) =0,at the same time, xi (T) is smooth and continuous at the moment T, and can obtain +.>Returning the system track to the specified track at the moment of the finite time T;
at this time, the track tracking error and the speed tracking error are respectively
Wherein e 1 (t)＝[e 11 (t)，e 12 (t)，e 13 (t)] T ，e 1g (t) track tracking errors in x-axis, y-axis and rotation angle directions, respectively; e, e 2 (t)＝[e 21 (t)，e 22 (t)，e 23 (t)] T ，e 2g (t) the speeds in the x-axis, y-axis and rotation angle directions respectivelyA degree tracking error;
design auxiliary variableWherein z is 1 (t)＝[z 11 (t),z 12 (t),z 13 (t)] T ，z 1g (t) auxiliary variable errors in x-axis, y-axis and rotation angle directions, respectively; k (K) 1 ＝diag(k 11 ,k 12 ,k 13 )，Sig(ξ) α ＝[|ξ 1 | α sgn(ξ 1 ),...,|ξ n | α sgn(ξ n )] T ，/>0＜α＜1；
Enabling electromechanical system uncertainty estimationWeight matrix->Is beta * And (2) and
thus, the weight error is obtainedThe self-adaptive rate of the design weight is as follows
Wherein Γ and K 4 Is an adaptive rate parameter;
the tracking error system obtainable from equation (3) and equation (6) is:
the design finite time controller is as follows:
wherein the method comprises the steps ofK 2 And k 3 Is a controller parameter;
the lyapunov function was established as follows:
deriving (10) along the error system (8)
Substituting the finite time controller (9) into the formula (11) to obtain
Wherein ε=w (t) - β *T G (x (t)) taking the parameter k 3 > |ε| and substituting the adaptive rate (7) into equation (11) to obtain
According to formula (12), there are further:
thus, from the formula (12)Wherein->μ＝(1+α)/2，1/2＜μ＜1，/>
Therefore, according to the finite time stabilization theory, the robot can start from any position and the tracking error system can be stabilized for finite time, the rehabilitation walking training robot can rapidly track the motion trail appointed by the doctor in finite time, and meanwhile, the rapid finite adjustment time can be obtained
2. The method for controlling the rehabilitation robot to quickly and limitedly control based on the random configuration network man-machine uncertain model according to claim 1, wherein the method is based on the MSP430 series singlechip to outputThe PWM signal is provided for the motor driving unit, so that the robot can track the motion track and the speed of the reference track signal simultaneously, an MSP430 series singlechip is used as a main controller, and the input of the main controller is connected with the motor speed measuring module and the output of the main controller is connected with the motor driving module; the motor driving module is connected with the direct current motor; the power supply system supplies power to each electrical device; the control method of the main controller is to read the feedback signal of the motor encoder and the control command signal given by the main controllerAnd->Calculating to obtain an error signal; according to the error signal, the main controller calculates the control quantity of the motor according to a preset control algorithm and sends the control quantity to the motor driving module, and the motor rotates to drive the wheels to maintain self balance and move in a specified mode.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011357061.5A CN112433495B (en) | 2020-11-27 | 2020-11-27 | Rehabilitation robot rapid finite time control based on SCN man-machine uncertain model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011357061.5A CN112433495B (en) | 2020-11-27 | 2020-11-27 | Rehabilitation robot rapid finite time control based on SCN man-machine uncertain model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112433495A CN112433495A (en) | 2021-03-02 |
| CN112433495B true CN112433495B (en) | 2023-11-21 |
Family
ID=74699331
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011357061.5A Active CN112433495B (en) | 2020-11-27 | 2020-11-27 | Rehabilitation robot rapid finite time control based on SCN man-machine uncertain model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112433495B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113359470B (en) * | 2021-07-05 | 2023-08-11 | 沈阳工业大学 | Designated transient time stability control method for restraining structural change of rehabilitation robot |
| CN113359767B (en) * | 2021-07-05 | 2023-08-18 | 沈阳工业大学 | Method for controlling safe driving of limited track tracking error of robot structure with slow change |
| CN113419423B (en) * | 2021-07-05 | 2023-10-24 | 沈阳工业大学 | Tracking control method for service robot to adapt to structural change in limited time |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE3028608A1 (en) * | 1980-07-28 | 1982-02-11 | Bernd Ing. Krieg (grad.), 2000 Hamburg | Hydraulically controlled artificial leg - has ram behind single axis knee joint with annular chamber containing valves |
| US4833381A (en) * | 1987-12-31 | 1989-05-23 | Westinghouse Electric Corp. | Optical automatic seam tracker and real time control system for an industrial robot |
| EP0323277A2 (en) * | 1987-12-31 | 1989-07-05 | Westinghouse Electric Corporation | Parametric path modeling for an optical automatic seam tracker and real time robotic control system |
| US5371669A (en) * | 1992-06-18 | 1994-12-06 | The United States Of America As Represented By The Administrator Of The National Aeronautics And Space Administration | Sliding mode control method having terminal convergence in finite time |
| US5455763A (en) * | 1992-12-30 | 1995-10-03 | Framatome | Process control method and device wherein models of the process and the control system are used to correct an input set point signal |
| CN104635738A (en) * | 2014-11-21 | 2015-05-20 | 沈阳工业大学 | Precise trace tracking optimal control method for uncertain rehabilitation walking training robot |
| CN107479381A (en) * | 2017-08-29 | 2017-12-15 | 沈阳工业大学 | Each axle tracking error optimal preventive control method of redundancy rehabilitation ambulation training robot |
| CN110989589A (en) * | 2019-11-30 | 2020-04-10 | 沈阳工业大学 | Tracking control method for rehabilitation walking robot with different trainers with randomly-changed mass |
-
2020
- 2020-11-27 CN CN202011357061.5A patent/CN112433495B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE3028608A1 (en) * | 1980-07-28 | 1982-02-11 | Bernd Ing. Krieg (grad.), 2000 Hamburg | Hydraulically controlled artificial leg - has ram behind single axis knee joint with annular chamber containing valves |
| US4833381A (en) * | 1987-12-31 | 1989-05-23 | Westinghouse Electric Corp. | Optical automatic seam tracker and real time control system for an industrial robot |
| EP0323277A2 (en) * | 1987-12-31 | 1989-07-05 | Westinghouse Electric Corporation | Parametric path modeling for an optical automatic seam tracker and real time robotic control system |
| US5371669A (en) * | 1992-06-18 | 1994-12-06 | The United States Of America As Represented By The Administrator Of The National Aeronautics And Space Administration | Sliding mode control method having terminal convergence in finite time |
| US5455763A (en) * | 1992-12-30 | 1995-10-03 | Framatome | Process control method and device wherein models of the process and the control system are used to correct an input set point signal |
| CN104635738A (en) * | 2014-11-21 | 2015-05-20 | 沈阳工业大学 | Precise trace tracking optimal control method for uncertain rehabilitation walking training robot |
| CN107479381A (en) * | 2017-08-29 | 2017-12-15 | 沈阳工业大学 | Each axle tracking error optimal preventive control method of redundancy rehabilitation ambulation training robot |
| CN110989589A (en) * | 2019-11-30 | 2020-04-10 | 沈阳工业大学 | Tracking control method for rehabilitation walking robot with different trainers with randomly-changed mass |
Non-Patent Citations (3)
| Title |
|---|
| Safe Exploration: Addressing Various Uncertainty Levels in Human Robot Interactions;Liu, CL等;2015 AMERICAN CONTROL CONFERENCE;465-470 * |
| 不确定康复训练机器人速度与加速度同时约束的跟踪控制;孙平;北京理工大学学报;第38卷(第10期);1067-1072 * |
| 不确定轮式移动机器人的有限时间镇定控制;叶锦华等;华南理工大学学报( 自然科学版);第41卷(第8期);48-54 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112433495A (en) | 2021-03-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112433495B (en) | Rehabilitation robot rapid finite time control based on SCN man-machine uncertain model | |
| Ijspeert et al. | Trajectory formation for imitation with nonlinear dynamical systems | |
| CN109465825A (en) | The adaptive dynamic surface control method of the RBF neural of mechanical arm flexible joint | |
| CN107479381B (en) | Optimal prediction control method for tracking error of each axis of redundant rehabilitation walking training robot | |
| CN110989589B (en) | Tracking control method for rehabilitation walking robot with different trainers with randomly changed quality | |
| CN112506054B (en) | Rehabilitation robot random finite time stable control based on SCN observation active thrust | |
| CN104698848A (en) | Control method for rehabilitation training of lower extremity exoskeleton rehabilitation robot | |
| CN103279039A (en) | Robot neural network type computed torque controller training platform and training method | |
| CN107831667B (en) | Tracking control method for compensating human-computer interaction force of rehabilitation walking training robot | |
| Yin et al. | Personalised control of robotic ankle exoskeleton through experience-based adaptive fuzzy inference | |
| Li et al. | Force sensing and compliance control for a cable-driven redundant manipulator | |
| CN112433475B (en) | SCN system offset identification-based cushion robot time-limited learning control method | |
| CN113359767B (en) | Method for controlling safe driving of limited track tracking error of robot structure with slow change | |
| Guo et al. | The robot arm control based on rbf with incremental pid and sliding mode robustness | |
| CN112571424A (en) | Direct constraint control of each axis speed of rehabilitation robot based on SCN walking force estimation | |
| Ayten et al. | Real-time trajectory tracking control for electric-powered wheelchairs using model-based multivariable sliding mode control | |
| CN113325720A (en) | Self-adaptive tracking control method for rehabilitation training robot with movement speed decision | |
| Jiang | Trajectory control of robot manipulators using a neural network controller | |
| CN113419423B (en) | Tracking control method for service robot to adapt to structural change in limited time | |
| CN116688453A (en) | Control method for direct switching active soft training of passive tracking training of rehabilitation robot | |
| CN116000917A (en) | Motion trail safety triggering data driving control method of rehabilitation walking robot | |
| CN112433474A (en) | Safety triggering control method of cushion robot based on SCN internal interference force estimation | |
| CN114859727A (en) | Rehabilitation robot preset learning time iterative control method based on speed observation | |
| Chen et al. | Adaptive robust constraint-following control for lower limbs rehabilitation robot | |
| Jiang et al. | Trajectory tracking control of industrial robot manipulators using a neural network controller |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |